Determinants of Emotion Work

Determinants of Emotion Work

Dissertation

zur Erlangung des Doktorgrades

der Mathematisch-Naturwissenschaftlichen Fakultäten der Georg-August-Universität zu Göttingen

vorgelegt von

Andrea Fischbach

aus Montabaur

Göttingen 2003

D 7

Referent: Prof. Dr. G. Lüer Korreferent: Prof. Dr. U. Lass Tag der mündlichen Prüfung:

Acknowledgments

I would like to thank Prof. Dr. Gerd Lüer and Prof. Dr. Uta Lass for supervising this dissertation and their helpful comments that helped develop this paper into its present version.

Special thanks are due to Prof. Dr. Dieter Zapf who encouraged my research in emotion work, shared his ideas, and provided support throughout many problems. Discussing the subject with him was always fruitful and inspiring.

I would also like to thank the many students who were involved in collecting data and developing research ideas: first, my students from the experimental courses in the last four years; and second, Dipl.-Psych. Kerstin Kielhorn, Dipl.-Psych. Katrin Meyer, Dipl.-Psych. Juliane Retzlaff, and Dipl.- Psych. Astrid Selke who did their theses on emotion work and helped collect data for this dissertation as well as shared their ideas. Research with these students was always fun and I can not imagine this dissertation without their help.

Margarita Neff-Heinrich assisted in the preparation of this dissertation by proofreading and offering helpful suggestions for improving its readability. She was also a real friend in stressful times and available any time I needed her—

motivating me with her "we will make it" in times I couldn't imagine it at all.

Thank you.

And lastly, a heartfelt "thank you" to my men on the home front—husband Werner and two boys Till and Falk—who spent many weekends without me surviving on more pizza dinners than usual—which probably bothered me more than them.

List of Tables and Figures IV

Research Note VI

1 Emotion Work 1

2 The Redefinition Self-Regulation Model of

Emotion Work (RS Model) 6

2.1 The Hackman framework for assessing the effects

of tasks 7

2.2 Emotion work external tasks 10

2.3 Emotion work task redefinition process 20

2.4 Strategies of emotion work 28

2.5 Consequences of emotion work 33

2.6 Summary of the RS Model 35

2.7 Summary of propositions 37

3 External Task Determinants of Emotion Work 40

3.1 Field Study Part 1Occupational Determinants of

External Emotion Work Tasks 49

3.1.1 Method—Field Study Part 1 56

3.1.2 Results—Field Study Part 1 62

3.1.3 Discussion—Field Study Part 1 65

3.2 Experiment 1 68

3.2.1 Method—Experiment 1 69

3.2.2 Results—Experiment 1 73

3.2.3 Discussion—Experiment 1 75

3.3 Experiment 2 77

3.3.1 Method—Experiment 2 78

3.3.2 Results—Experiment 2 81

3.3.3 Discussion—Experiment 2 83

3.4 Experiment 3 84

3.4.1 MethodExperiment 3 86

3.4.2 Results—Experiment 3 88

3.4.3 Discussion—Experiment 3 90

4 Redefinition Determinants of Emotion

WorkField Study Part 2 93

4.1 MethodField Study Part 2 95

4.2 ResultsField Study Part 2 105

4.3 DiscussionField Study Part 2 112

5 General Discussion 116 5.1 How study findings relate to other emotion work

research 116

5.2 Strength and limitations of the study 125

5.3 Job Design in Emotion Work 128

Summary 136

References 140

Curriculum Vitae 148

Tables

Table 1 Main Occupational Differences 52

Table 2 Correlations among Study Variables and

Descriptive Statistics (Field Study Part 1) 62

Table 3 Multiple Mean Comparisons of FEWS 64

Table 4 Professional Identity Scales 103

Table 5 Correlations among Study Variables and

Descriptive Statistics (Field Study Part 2) 106

Table 6 Organizational Socialization and Professional

Identity as Determinants of Emotion Work 107

Figures

Figure 1 Hackman Framework for Assessing the Effects of Tasks 9 Figure 2 Redefinition Self-Regulation Model of Emotion

Work (RS Model) 37

Figure 3 First part of the RS Model 41

Figure 4 Simplified RS Model: Effects of External Tasks

on Self-Regulation, Task Behavior, and Consequences 48

Figure 5 Job profiles in Emotion Work Scales 64

Figure 6 Effect of Display Rule on Effort (Experiment 1) 74 Figure 7 Effect of Display Rule on State Negative Affect

(Experiment 1) 75

Figure 8 Effect of Display Rule on Emotional Dissonance

(Experiment 2) 82

Figure 9 Effect of Display Rule on State Negative Affect

(Experiment 2) 83

Figure 10 Effect of Display Rule on Emotional Dissonance

(Experiment 3) 89

Figure 11 Effect of Display Rule on SERVQUAL

(Experiment 3) 90

Figure 12 Simplified RS Model: Effects of External Tasks

and Personal Determinants on Task Redefinition 94

In Field Study Part 1 and Field Study Part 2, teachers and travel agents data were surveyed by students as part of their thesis. The reported experiments were accomplished by students within three practical courses for course credits, supervised by the author of this dissertation.

Results of Experiments and Field Study Part 1 and 2 were presented at several conferences.

1 Emotion Work

Imagine that you are a police officer policing a demonstration. Suddenly a demonstrator spits on you. How would you instinctively behave in this situation?

How should you behave in this situation? Would these behaviors match?

Typically there are implicit or explicit organizational expectations regarding how to behave emotionally, such as police officers remaining neutral and objective when interacting with citizens (Fischbach, 2000; Fischbach & Zapf, 2002).

Emotion work (emotional labor) can be defined as emotional job demands and psychological strategies necessary to regulate these demands. Emotional job demands can be summarized into five requirements and one stressor. These five job requirements are the display of positive emotions, negative emotions, neutrality, sensitivity, and sympathy. The job stressor, emotional dissonance, is demanded when these emotional job requirements are not actually felt. This definition allows for a broad, inclusive understanding of emotion work from an organizational perspective, extending beyond efforts and labor limited to the worker.

This study focuses on how emotion work can be understood as a job demand—particularly how emotional dissonance can be understood as a job stressor, and distinguishes emotional demands from strategies applied to fulfill

these job demands as well as from the consequences of emotion work. Several authors discuss aspects of emotion work concepts as job demands (Brotheridge

& Lee, 1998; Grandey, 2000; Zapf, Vogt, Seifert, Mertini, & Isic, 1999) but questions remain regarding how to distinguish emotion work job demands from individual work behavior. Controversial views exist, especially regarding the job stressor, emotional dissonance, which was proposed from the onset of emotion work research to affect negative individual consequences (Abraham, 1998;

Brotheridge & Grandey, 2002; Hochschild, 1983; Morris & Feldman, 1996, 1997; Zapf, Seifert, Schmutte, Mertini, & Holz, 2001). The innate feelings of the police officer in the above situation probably do not match the organizationally- required neutrality, requiring the officer to master (ignore or suppress) an inner conflict in order to comply with his or her job demands. Such situations highlight whether emotional dissonance is limited to a question of individual strategies of self-regulation (Bandura, 1977, 1986; Kanfer, 1990), what emotions are individually triggered by the work situation, what kind of emotional task behavior is displayed, and to what extent this behavior is in line with inner feelings. In contrast to the organizational-focused perspective of this study, some emotion work research takes a more worker-focused perspective regarding emotional dissonance and its negative consequences on individual well-being.

Studies that are worker-focused (Brotheridge & Grandey, 2002; Brotheridge &

Lee, 1998, 2002; Totterdell & Holman, 2003) consider emotional dissonance as a self-regulation strategy with regard to affective events theory (Weiss &

Cropanzano, 1996). These studies propose that emotions triggered by situations such as spitting at a police officer have to be regulated by hiding, faking, and/or modifying feelings—leading to different, individual-determined consequences.

Studies that are organizational-focused consider emotional dissonance an organizational job demand independent of individual strategies to regulate this demand—a job stressor leading to predictable negative consequences (Zapf, 2002; Zapf et al., 2001).

This study is based on emotion work literature, theoretical concepts, and empirical findings to date (for an overview see Zapf, 2002) and draws from Hackman's (1969, 1970) concept of work behavior and his theoretical framework for studying consequences of organizational tasks. In chapter 2, several conceptualizations and perspectives in emotion work research are integrated into a framework entitled Redefinition Self-Regulation Model of Emotion Work (RS Model). The RS Model attempts to distinguish emotion work demands (job requirements and the stressor emotional dissonance) from strategies in dealing with these job demands and their consequences. In emotion work research, this study is the first to investigate the role of redefinition processes specific to emotion work, redefinition being the cognitive processes involved in transforming an external-assigned task into one's understanding of what one ought to do (Hackman). The RS Model illustrates how this emotion work conceptualization is in concert with general work psychological perspectives.

Within the framework of integrated conceptualizations and perspectives in emotion work research, several propositions are deduced from the RS Model about both the antecedents as well as the consequences of emotion work demands and strategies. In an attempt to demonstrate the validity of the majority of these propositions, three pivotal research topics in understanding emotion work are derived from the RS Model and empirically tested in this study: (1) organizational variables affect the definition of emotion work job demands, (2)

emotional dissonance is a stressor in service work, affected by organizational variables as any other job demand would be, and (3) socialization strategies as an organizational determinant and professional identity as a personal determinant are important variables in understanding the redefinition of emotion work job demands. The first and second research topics are covered in chapter 3External Task Determinants (external task (Hackman) refers to the external-assigned aspect of emotion work job demands). For the first research topic regarding organizational variables defining emotion work demands, a field study of three highly diverse occupations (police, teachers, and travel agents) is applied to show predictable differences in emotion work job requirements and the stressor, emotional dissonance. The hypotheses regard the differences in the major tasks (organizational-defining tasks) and primary tasks (goal-defining tasks) of these three occupations, which include differences in service- and interdependence characteristics (e.g., customer/client demands, status of service workers, customer/client control parameters). The second research topic—emotional dissonance as a stressor in service work—includes an experimental paradigm to test the proposition that emotional dissonance as an external task is, among the other variables, particularly defined by organizational display rules (organizational expectations regarding emotions; Ekman & Friesen, 1975) and customer event intensities and qualities (Morris & Feldman, 1996). Given intensely rude customer behavior where an average person spontaneously feels disgust or anger, emotional dissonance should predictably be evoked by the experimental display rule to be friendly (customer oriented in all circumstances);

whereas the display rule to be authentic (be natural and show your true

personality in all circumstances) should evoke less emotional dissonance. Three experiments are applied to test this experimental paradigm and hypothesized consequences of display rule manipulation. The third research topic regarding socialization strategies and professional identity is the subject of chapter 4Redefinition Determinants (redefinition (Hackman) refers to the individual- understanding aspect of emotion work job demands). The predicted effects of socialization strategies (how organizations communicate their display rules) and professional identity (self-imposed professional role expectations) on reported emotion work requirements and emotional dissonance in police, teacher and travel agent samples are analyzed by means of hierarchical regression analysis.

In chapter 5, the findings of this study are discussed with respect to conclusions and suggestions regarding emotion work job design/redesign and management.

In chapter 6, the entire study is summarized.

2 The Redefinition Self-Regulation Model of Emotion Work (RS Model)

In 1983, Arlie Russell Hochschild published the seminal emotion work/emotional labor book, The Managed Heart. Commercialization of Human Feeling. The emotion work research this book set in motion differs in focus regarding how emotion work (a) is defined and identified as a job requirement and/or job stressor, (b) is coped with (strategies) by employees, and/or (c) impacts organizational consequences. The integrative framework of this study is based on these three themes of emotion work and stress literature and models of antecedents and consequences of emotion work (Grandey, 2000; Schaubroeck &

Jones, 2000; Zapf, 2002) as well as Hackman's (1969, 1970) seminal framework for assessing stress effects of tasks. The Redefinition Self-Regulation Model of Emotion Work (RS Model) presented in this study is an attempt to integrate diverse conceptualizations and perspectives in emotion work. The RS Model is not intended to be a formal model (direct hypotheses-testing model) of emotion work nor inclusive of every emotion work principle and nuance. The intention of the RS Model is to present a possibly useful understanding of emotion work aspects, sorting out variables which might be important in understanding emotion work antecedents and consequences, applying findings of emotion work research

to these variables, and deriving new propositions and hypotheses about antecedents and consequences of emotion work.

2.1 The Hackman framework for assessing the effects of tasks

The Hackman (1969, 1970) framework for assessing the effects of tasks differentiates (a) the input-stagewhere an external task is perceived and interpreted by individuals (redefinition process) into an internal task; which in turn affects, (b) the strategy-stageincludes internal processes (personal approaches to carrying out a task and personal evaluations of trial outcomes) as well as actual task behavior; which in turn affects (c) the final outcome- stageconsequences of the Hackman-proposed task execution process. Figure 1 (p. 9) is a simplified illustration of this three-stage framework presented by Hackman (1970, p. 213). For purposes of the present studyto show how emotion work might be understood as a job demand and particularly how emotional dissonance might be understood as a stressorthe focus of Figure 1 is on the input-stage elements (external task, task redefinition process, and internal task) and salient constitutional and influencing variables of these elements.

Consequently, Hackman's (1969, 1970) multiple effects of personal factors in the strategy stage as well as the outcome stage are simplified.

Task definition. External task is defined by Hackman as: "… assigned to a person (or group) by an external agent or is self-generated, and consists of a stimulus complex and a set of instructions which specify what is to be done vis- à-vis the stimuli. The instructions indicate what operations are to be performed by the performer(s) with respect to the stimuli and/or what goal is to be

achieved." (Hackman, 1970; p. 210). According to Hackman (1969, 1970), external agent/self-generated determinants affecting the external task are goal instructions, operation instructions, and stimulus material. This definition serves the organizational focus vs. worker focus of the RS Model as it defines the sources of external tasks as being independent of individual work behavior, making it possible to systematically vary objective parameters in experimental situations and investigate their consequences.

Task redefinition process. The second element of Hackman's framework is related to his proposition that an external task is perceived and interpreted by the task performer, who cognitively redefines the task ("what ought I to do?").

Hackman called this process task redefinition, and the outcome variable of this task redefinition, internal task. According to Hackman, personal determinants affecting the task redefinition process involve the individual's understanding of the external task (either adequately or inadequately); acceptance of the task;

idiosyncratic needs, values, attitudes, etc.; previous experience with the same or similar tasks; ability; performance motivation; and level of arousal.

Consideration of these personal determinants may help to explain why the same external task is often represented and executed differently by different people.

Once again this Hackman definition serves the RS Model as it makes clear that the same external task can lead to differences in internal tasks with respect to personal determinants, making it possible to investigate both external task input and personal determinant input and identify their consequences.

Task performance strategies and outcomes. According to Hackman, the input-stage (external tasks redefined into internal tasks) leads to cognitive processes of the task performer, who develops personal approaches relevant to

the strategy of performance and relevant to actual behavior (internal processes).

This leads to actual task behavior strategies, which in turn has trial outcomes, which in turn affect new strategies of performance (internal processes). This process leads to certain outcomes (consequences). Once again this serves the RS Model as it makes clear that the strategies (internal cognitive processes and task behavior) affect work outcome variables (consequences), making it possible to investigate these strategies and identifying their consequences.

Input ^{Strategy Outcome}

External

Task Internal

Task

Redefinition ^Task

Behavior Conse-

quences Internal

Process

Personal Determinants

− Understanding of task

− Acceptance of task External agent/

self-generated − Idiosyncratic needs, values, attitudes, etc.

− Previous experience with similar tasks

− Goal instructions

− Operation

instructions − Ability

− Performance motivation

− Stimulus material

− Level of arousal

Figure 1. Hackman Framework for Assessing the Effects of Tasks (1970, p.213).

2.2 Emotion work external tasks

Emotion work tasks and service. Emotion work occurs in jobs where service- worker interactions with customer/clients are required. In service occupations, there are explicit and/or implicit organizational expectations regarding management of emotions, that is: how one's own emotions and the emotions of the interaction partners (customers/clients) should be handled. Hochschild (1983) suggests that there are organizational display rules that specify which emotions (if any) to express in each type of customer interaction. The Hochschild conceptualization is that emotion work is an organizational job demand specific to people work, where emotions have to be displayed and managed by employees in an organizationally-defined manner. An application of this concept is where, in order to close a sale, a salesperson displays friendliness and concern or a doctor might calm a child by joking during the examination (Nerdinger, 2001).

It is important to note that emotion work is not equivalent to service work—there are employees of service occupations without customer/client contact and there are occupations which are not considered service occupations in the economic science definition of service occupations but nevertheless require customer/client interactions (Nerdinger, 1994; Zapf, 2002). The current study applies a psychological interpretation of service (referred to as people work or person-related tasks in emotion work literature) and therefore a psychological use of the terms service organization as well as service worker. In this study, police and teachers are referred to as service occupations even though from an economic science point of view neither police nor teachers qualify as service sector occupations (Nerdinger, 1994). In economic science, a customer or client

who requests and pays for service is fundamental to service industry definitions—yet in neither of these occupations is a customer paying for service (police and teachers are paid by the society). It is also most likely that under many circumstances, the recipients of police and teacher service do not request said service (e.g., being cited for exceeding speed limit; learning differential math equations). However, police as well as schools are characterizing themselves more and more as service organizations. From an emotion work point of view, these occupations consist of relevant person-related tasks and service- and interdependence characteristics.

Determinants of external tasks in emotion work. According to Hackman (1969, 1970), an external task can be defined either by an external agent or self generated and is greatly influenced by goal instructions, operation instructions, and stimulus material. In emotion work, the external agent that defines external tasks is the service organization (organizations engaged in people work) with the attendant goal- and operation instructions, and stimulus material (termed organizational determinants in the RS Model). Hochschild has set in motion a body of psychological conceptualizations (Ashforth & Humphrey, 1993;

Grandey, 2000; Morris & Feldman, 1996; Nerdinger, 1994; Zapf, 2002) and empirical research on emotion work (Abraham, 1998; Brotheridge & Grandey, 2002; Kruml & Geddes, 2000; Schaubroeck & Jones, 2000; Zapf et al., 1999;

Zapf et al., 2001; Zerbe, 2000). Summarizing these emotion work conceptualizations and empirical findings relative to external tasks, the framework proposed in this study applies the Hackman perspective of external tasks involving multiple determinants, organizing the emotion work external task determinants under the heading of organizational determinants with the three

categories of: (a) occupational determinants relative to the type of service organization, (b) inner organizational determinants relative to organization- specific attributes, and (c) customer event determinants relative to service- worker-customer/client interactions (Figure 2, p. 37, RS Model).

Occupational determinants of external tasks in emotion work. Based on the major tasks and adherent major goals of the service organization, object- related primary tasks (Rice, 1963) and person-related primary tasks and/or subtasks can be derived. Organizational-defining major tasks differ according to occupations. In sales services, a major task might be to sell a certain product and to reach the adherent goals. Object-related primary tasks (e.g., purchasing, stock keeping) are derived and person-related subtasks (e.g., describing aspects of products to customers, giving customers advice) are derived from primary tasks.

For doctors, a major task would be health care and a derived person-related primary task would be the diagnosis of a patient. To reach the goals of person- related primary tasks (e.g., acquire a patient's medical background), social interaction subtasks also have to be achieved (Strauss, Farahaugh, Suczek, &

Wiener, 1980; Zapf, 2002). The present study presumes, in accordance with Zapf (2002), that emotion work is a subtask of a person-related task. Qualitative studies have shown that service occupations have discrete emotional job demands according to occupational-specific tasks (Rafaeli, 1989a; 1989b). The organizational major tasks relate to the Hackman concept that goal instructions define the external task (Figure 1, p. 9). This study postulates that occupational differences in major tasks affect the definition of external tasks in emotion work.

In psychological approaches to service research, there are three well acknowledged service characteristics: intangibility, heterogeneity and

inseparability (Bowen & Schneider, 1988; Maister, 1982; Nerdinger, 1994;

Parasuraman, Zeithaml, & Berry, 1988; Schneider & Bowen, 1984). Intangibility pertains to the issue that service is a construct and that the quality of service can be evaluated only by subjective experiences. Heterogeneity relates to the issue that service encounters are never identical with different customers or even with the same customer at different times. Inseparability describes the phenomenon whereby the production and consumption of service occurs simultaneously.

Occupations can differ in the complexity of these service characteristics: the quality of service in a restaurant is easier to evaluate than the quality of a medical diagnosis (Zeithaml, 1981, citation according to Nerdinger, 1994). Heterogeneity is most likely less intense in a shop where single, short interactions between strangers follow behavioral scripts than in a hospital with patients remaining an entire week or longer. Inseparability is typically high for direct person-related service (human services) where the customer/client is the direct object of service (e.g., services provided by nurses, physicians, teachers, and social workers) and low for indirect person-related service where something pertaining to the customer/client is the object of service (e.g., services provided by repair persons, salespersons, tellers), or the arrangement of consumption of a product (e.g., services provided by insurance administrators, investment brokers) (Gross, 1983;

Nerdinger, 1994). The organizational service characteristics relate to the Hackman concept that the stimulus material defines the external task (Figure 1, p. 9). This study postulates that occupational differences in service characteristics affect the definition of external tasks in emotion work.

There are many typologies of service occupations (Nerdinger, 1994). One considers the nature and intensity of interdependence between the service worker

and customer/client. Interdependence involves the dynamic of service workers expressing emotions to influence customer/clients (Kruml & Geddes, 2000), customer/clients doing the same to influence service workers (Rafaeli, 1989a, 1989b), and that in different occupations, influence strategies and influence possibilities can differ on both sides of the dynamic (Nerdinger, 1994). The first of three aspects of this interdependence concept is the quality of the interaction.

Based on Gutek (1997), Holman (2003) classified the quality of service interactions in call centers as encounters vs. relationships. Encounters can be characterized as single, short interactions between strangers, following behavioral scripts where both interaction partners are presumed to not show authentic emotional expression. Service interactions classified as relationships are where the interaction partners typically know each other, have a shared history, and interactions have relatively more leeway where both interaction partners are presumed to show authentic feelings involving trust and loyalty.

Zapf, Isic, Bechtoldt, and Blau (2003) argue according to Gutek (1997) and Holman (2003) that direct person-related services can be classified more as relationship interactions whereas indirect person-related services can be described more as encounter interactions, and that different emotion work display rules can be expected according to the different qualities of interactions.

The second aspect of interdependence is customer demands. For most service organizations, customer demands are essential to defining organizational goals.

For instance, hospital patients expect service based more on the relationship interaction (e.g., the nurse calls him/her by name and shows sympathy); whereas a fast food customer most likely expects service according to the encounter interaction (e.g., standardized greeting asking for food order). Rafaeli and Sutton

(1990) showed in a structured observation study how customer demands affect actual display of positive emotions of supermarket cashiers (e.g., eye contact, smiling, pleasantness, thank-yous). In sales and other service organizations, high standards in service quality and high customer satisfaction are often explicit goals (Zeithaml & Bitner, 2000). The third aspect of interdependence is the status and control of service workers relative to customer/clients (Nerdinger, 1994). Doctors have high status and interaction control compared to their patients. It can be assumed that in the context of this interdependence aspect, emotion work demands of doctors and other high status/high interaction control professionals differ from a cashier in a supermarket (Rafaeli, 1989a). The organizational interdependence characteristics relate to the Hackman concept that the stimulus material defines the external task (Figure 1, p. 9). This study postulates that occupational differences in interdependence characteristics affect the definition of external tasks in emotion work.

Taken together, it can be proposed that an external task in emotion work is determined by the occupational major tasks, service characteristics, and interdependence characteristics (Proposition 1).

Inner organizational determinants of external tasks in emotion work.

During face-to-face or voice-to-voice interactions, service workers are required to express appropriate emotions as a job demand to help achieve organizational goals and enhance the bottom line. In line with the Goffman (1959) role concept of social interaction, Hochschild (1983) used the metaphor of a drama—where the work setting is seen as a stage, the customer/client is the audience, and the employee is the actor with explicit and/or implicit display rules governing how to

interact with the customer/client (Ekman & Friesen, 1975). The concept of display rules exists in much of the emotion work literature. This study relates display rules in emotion work to the Hackman variables goal- and operation instructions (Figure 1, p. 9). Flight attendants are generally required to show positive emotions (Hochschild, 1983); bill collectors (Sutton, 1991) and detectives (Stenross & Kleinman, 1989) are required to show situational- determined negative, positive or neutral emotions; and nurses are particularly required to show sympathy (Buessing & Glaser, 1999; James, 1989, 1992). There is a body of literature that deals with the question of how, in different corporate cultures, display rules are communicated and how service workers are influenced by various socialization strategies (e.g., Ashforth & Humphrey, 1993;

Hochschild, 1983; Rafaeli, 1989a, 1989b; Rafaeli & Sutton, 1987; Van Maanen

& Kunda, 1989). Sutton (1991) describes how organizations use socialization strategies such as personnel selection, written materials, supervisor and peer observations, training, and rewards and punishments to manage the emotion norms regarding their bill collector interactions with debtors. Tsai (2001) showed that the psychological climate for service friendliness in a corporate culture was positively related to positive emotions displayed by service workers. This study relates socialization strategies and corporate culture in emotion work to the Hackman variable stimulus material (Figure 1, p. 9). These conceptualizations—

display rules, socialization strategies, and corporate culture—are listed as inner- organizational variables in the RS Model (Figure 2, p. 37).

Taken together, it is proposed that an external task in emotion work is determined by inner organizational display rules, socialization strategies, and corporate culture (Proposition 2).

Customer event determinants of external tasks in emotion work. Most emotion work empirical studies consider the frequency of emotional display as an important aspect of emotion work (Adelmann, 1995; Brotheridge & Lee, 1998, 2002; Morris & Feldman, 1996, 1997; Zapf et al., 1999). When a call center agent speaks with 60 to 250 customers per eight-hour shift (Dieckhoff, Freigang-Bauer, Schröter, & Viereck, 2002), these high numbers of service encounters can be expected to lead to higher frequencies of emotional displays compared to a therapist who might treat a maximum of eight patients per day. It is also plausible that the intensity and duration of emotional display requirements are further aspects of emotion work, suggested by Morris and Feldman (1996) to require attentiveness to display rules. Regarding the quality of customer events, call center agents often report interactions with rude and impolite customers (Dormann, Zapf, & Isic, 2002), whereas therapists presumably have less negative interactions with clients. The required variety of emotions to be expressed can also differ. A call center agent is most likely expected to express only favorable emotions, but a kindergarten teacher might be expected to express a high variety of emotions ranging from kindness to strictness, depending upon the situation.

These factors related to customer events—frequency, intensity, duration, quality, and variety—are listed as customer event parameter values in the RS Model.

This study relates customer events to the Hackman external task determinant, stimulus material.

Taken together, it is proposed that an external task in emotion work is determined by customer event frequency, intensity, duration, quality, and variety (Proposition 3).

Interrelations of organizational determinants. This summarization of the emotional work literature dealing with what in this study is termed occupational, inner organizational, and customer event determinants of external emotion work tasks are proposed to affect the definition of an actual task. Variables referred to as inner organizational determinants (display rules, socialization strategies, and corporate culture) are presumed to be more similar within and more dissimilar among service occupations. It is also presumed that previously discussed customer event parameter values can be expected to be affected by occupational and inner organizational variables as well.

Thus, customer event determinants of external emotion work tasks are expected to be more similar within an organization and more dissimilar among organizations and more similar within an occupation and more dissimilar among occupations; and inner organizational determinants of external emotion work tasks are expected to be more similar within an occupation and more dissimilar among occupations (Proposition 4).

Emotional Dissonance as an external task in emotion work. According to Proposition 1 regarding occupational determinants of external tasks, Proposition 2 regarding inner organizational determinants of emotion work, and Proposition 3 regarding customer event determinants of emotion work, an external task in emotion work is proposed to be defined by a combination of these three aspects and their parameter values (Figure 2, p. 37, RS Model). It is suggested in the literature that display rules requiring the display of specific emotions are associated with situational cues which determine the external emotion work task (Rafaeli & Sutton, 1987, 1990). Given the display rule to be friendly toward customers in a situation where a customer is complaining in a very rude manner

(a common work experience in call-center work—Dormann et al., 2002), or given the display rule in police work to be neutral and professional in a death situation (Fischbach, 2000; Fischbach & Zapf, 2002), it is probably true in both cases that the evoked feelings are not in line with the display rule demands of such situations. As previously mentioned in this study, emotional dissonance occurs when an employee is required to express emotions which are not genuinely felt in the particular situation (Hochschild, 1983). This could be thought of as a person-role conflict in which a person's innate emotional response is in conflict with role expectations regarding the display of emotions (Abraham, 1998; Rafaeli & Sutton, 1987). Some see emotional dissonance more as a dependent variable (Adelmann, 1995; Ashforth & Humphrey, 1993), others refer to it as occurring in the emotion work self-regulation process (Brotheridge &

Grandey, 2002; Brotheridge & Lee, 1998, 2002; Totterdell & Holman, 2003).

Zapf (2002) proposes with regard to action theory and stress research (Frese &

Zapf, 1988, 1994; Zapf et al., 1999) that emotional dissonance is a built-in aspect of the work role and job design, or more specifically a job stressor leading to predictable negative individual consequences (Hochschild, 1983). According to Zapf (2002), the aforementioned propositions regarding external task definition in emotion work, and the application of an external task definition with respect to the Hackman framework (1969, 1970), this study proposes that emotional dissonance can more comprehensively be defined as an external work task rather than a personal psychological reaction to emotional display or a personal behavioral strategy. It is contended by this study that as a job demand (stressor), emotional dissonance occurs as a predictable, external task in customer events where the service worker will most likely either feel an emotion which is not in

line with the emotion required to be expressed (organizational display rule) or feel nothing. As an external task, emotional dissonance is subject to the same determinants, processes, strategies, and consequences as any other external task (Figure 2, p. 37, RS Model).

Thus, Proposition 5 suggests that emotional dissonance is an external emotion work task (Proposition 5).

2.3 Emotion work task redefinition process

The RS Model (Figure 2, p. 37) illustrates the redefinition of external tasks as being affected by two sources of input: the emotion work external task and personal determinants. Of all the organizational determinant variables influencing an emotion work external task, it can be argued that display rules and socialization strategies probably play the most important roles. It is generally agreed in emotion work literature that there are expectations of the organization regarding how service workers should behave when interacting with customers/clients (Hochschild, 1983; Schneider & Bowen, 1984; Zapf, 2002). It is also generally agreed that, compared to manufacturing work, it is much more difficult in service occupations to define specific organizational tasks (external tasks) which can be unambiguously redefined into internal tasks. What exactly does it mean to display role- and situation-appropriate emotions as part of one's job? It is the exception from the rule if a company has documented, explicit display rules governing how to behave in interactions with customers (e.g., Ash, 1984). Display rules (listed as an inner organizational variable in the RS Model) are more often part of the organizational culture and are learned through explicit and implicit processes of organizational socialization (Bauer, Morrison, &

Callister, 1998). This complicated nature of communicating and governing service worker display rules limits the possibilities of an organization to directly control not only how tasks are carried out but also organizational and individual consequences, and leads to an increased relevance of indirect organizational control mechanisms as well as individual personal determinants in the employee task redefinition process (Bowen & Schneider, 1988; Schneider & Bowen, 1984). As mentioned in section 2.1 and illustrated in Figure 1 (p. 9), Hackman postulates an extensive list of personal determinants that affect the task redefinition process. Regarding emotion work, four of these personal determinants are considered especially relevant: (a) understanding of task—

considered to be mostly affected by socialization strategies; (b) acceptance of task—which in the context of emotion work is postulated to be mostly a question of individual professional identity; (c) ability—which in the context of emotion work is considered to be mostly a question of emotional competence; and (d) idiosyncratic needs, values, attitudes, etc.—which in the context of emotion work are expected to be mostly a question of personality factors and gender.

Socialization strategies as an organizational determinant. In accordance with the literature regarding behavior expectations (Hochschild, 1983; Schneider

& Bowen, 1984; Zapf, 2002), the challenges of defining and communicating these display rules (e.g., Ash, 1984), and the Hackman claim that external task understanding in the redefinition process is dependent upon the clarity of the task, it is postulated that the pivotal aspect regarding the success of display rules is if and how service workers learn them. In job roles involving customer interactions, organizational display rules are important as they are presumed to define what the appropriate and expected job behavior is for a particular job.

Constituent for social roles is that they are sometimes difficult to pin down because individual opinions can differ over what one's role should be (Scott, Mitchell, & Birnbaum, 1981). The professional role-making/role-taking process (Mead, 1934) is important to organizational socialization and can be mapped out into stages— beginning with group expectations for a particular position, communication processes about these expectations, perceived expectations about the role, actual role behavior, and the feedback of this role behavior compared to group expectations. The goal of socialization strategies is to adapt the employee values, attitudes and skills to the specific job requirement (Scott et al., 1981). It is assumed that the more socialization strategies an employee perceives, the more aware the employee will be regarding role expectations, and theoretically the greater influence the organization display rules will have in the task redefinition process regarding emotion work. Some examples set forth by the present study in which socialization strategies and sources of display rules can be defined and communicated explicitly are by: (1) explanations of superiors; (2) educational courses; (3) vocational training. Examples in which socialization strategies can be communicated implicitly are by: (4) perceived, unspoken part of the company culture; (5) individual job experiences; (6) individual beliefs about norms and values of a profession; and (7) perceived societal expectations of a profession which are typically communicated by the customer or client (Sutton, 1991). It can be contended that the more aware an individual is of explicit- or implicit- imparted role expectations (display rules) relative to their job, the more likely they are to redefine their imposed external tasks within the context of organizational expectations.

It is proposed in this study that socialization strategies as an organizational determinant of external emotion work tasks influence the redefinition process of emotion work (Proposition 6).

Professional identity as a personal determinant. In several studies, display rules are considered more or less conciliative to the individual professional ethos (Ashforth & Humphrey, 1993; Ashforth & Tomiuk, 2000;

Briner, 1995; Rafaeli & Sutton, 1987). In a qualitative study, Ashforth and Tomiuk revealed that the interviewed service agents generally believed that their professional task behavior reflect their true selves. They also generally reported that they had to be good actors and fake emotions toward customers/clients in order to do their job well. This seemingly contradictory phenomenon was discussed in the emotion work literature as internalized role playing (Hochschild, 1983; Morris & Feldman, 1996; Rafaeli & Sutton, 1987). Ashforth and Tomiuk described this phenomenon as deep authenticity where a required emotion (display rule) is in line with the display rules of a specific identity, internalized and displayed as a reflection of self regardless of whether genuinely felt. They related the concept of authenticity to the service worker question (more likely posed intuitively, holistically and automatic), "do I identify with the role?"—

professional identity. More clarity to the concept of professional identity can be derived from Stryker's Identity Theory (1987). Identity Theory is based on Symbolic Interactionism (Cooley, 1902; Mead, 1934), which postulates that individuals are members of several social categories (e.g., husband/wife, father/mother, sport-team-member/volunteer, police-officer, etc.). People learn role expectations and role demands primarily through interactions with others.

Each role produces distinct components of self, generally referred to as role identities. Role identities can usually be defined as self-conceptions, self-referent cognitions, or self-definitions according to the role(s) individuals occupy.

Identity, in turn, can usually be defined as the internalized set of these several role identities. Some aspects of a role identity have more self-relevance than others. The commitment to certain aspects of role identities (Foote, 1951) helps to explain the identity salience of that person. An important aspect of Stryker's Identity Theory is that identity seeks validation: the more salient an aspect of an identity, the more receptive one is to behavioral opportunities that confirm that aspect of identity. In service occupations, there are role expectations of the organization and of the customers. By interactions with members of the organization (management, staff, colleagues, and training groups) and with customers, service workers build a sense of who they are as a service professional—what their professional values, goals and beliefs are—and what ought to do and how they ought to behave on the job. According to Stryker, some aspects of the service role can become personally central, salient and valued (e.g., quality-service orientated); others can be discounted as personally unimportant and valueless (e.g., fast-service oriented). It is proposed that, in line with Stryker's Identity Theory, the most salient professional identity aspects of the individual will be evoked in service work situations, and are more likely to lead to redefinitions of external tasks into internal tasks that the individual considers important. It is postulated that awareness of the salient professional identity aspects of the service worker will aid in understanding how individuals redefine external emotion work tasks, and also aid in predicting consequences of emotion work tasks on said individuals. Again citing Stryker's Identity Theory,

within the same profession it can be expected that there are most likely a multitude of professional identity aspects, and that there are individual differences in the salience of these aspects. These individual differences in professional identity salience, in turn, lead to individual differences in the task redefinition process within the same occupation. The current study argues that in emotion work, professional identity (the internalized set of professional role expectations in one's job) is the fundamental aspect in task acceptance and relates to Hackman's acceptance of task personal determinant in task redefinition. The RS Model lists professional identity as the first of four personal determinants affecting the emotion work external task redefinition process.

It is proposed that professional identity as a personal determinant influences the redefinition process of emotion work (Proposition 7).

Emotional competence as a personal determinant. The Riggio (1986) model of basic social skills proposed that the perception of the emotions of interaction partners and the expressing and controlling of one's emotions in a nonverbal and verbal manner are fundamental social skills. The concept of emotional intelligence was conceptualized by Salovey and Mayer (1989) and popularized by Goleman (1996). Salovey and Mayer distinguished four psychological processes involving emotional information: (a) appraising; (b) expressing emotions in self and others; (c) using emotions adaptively to achieve one's goals; and (d) regulating emotions in self and others. The term competence rather than intelligence is preferred in this study because there are theoretical problems in calling these competences intelligence when distinguishing them from common psychological intelligence concepts (Davies, Stankov, & Roberts,

1998; Schuler, 2002). These competences (and the experiences in applying them in customer relations) affect the personal choice of which emotion display is demanded/appropriate in a given situation. A kindergarten teacher perceiving as naughtiness an outburst of fury from a child over an unsolved puzzle will redefine what emotion display is required in this situation more in the manner of strictness and punishment compared to a kindergarten teacher perceiving the outburst as frustration and will more likely redefine the task in a sympathetic, calming manner. Or the experience that children often calm down faster when outbursts are ignored rather than positively reinforced with attention might lead to other task redefinitions such as remaining neutral. This study proposes that in the context of personal determinants influencing the emotion work task redefinition process, emotional competence relates to the Hackman ability variable of personal determinants. The RS Model lists emotional competence as the second of four personal determinants affecting the emotion work task redefinition process.

It is proposed that emotional competence as a personal determinant influences the redefinition process of emotion work (Proposition 8).

Personality and gender as personal determinants. The emotion work literature and empirical research treats other important aspects which fit into the RS Model framework of affecting the redefinition of emotion work external tasks—specifically personality traits and gender (Grandey, 2000; Hochschild, 1983; Schaubroeck & Jones, 2000). The malleability of aspects affecting emotion work leads to what is termed a weak situation in personality research. It is proposed in this study that emotion work can be assumed to be an organizational

job demand where many weak situations governed by individual differences occur, and therefore inter-individual behavior differences should be detectible.

Schutte, Kenrick, and Sadalla (1985) found that weak situations are often associated with behavioral variance between individuals, one explanation being that individual personality differences influence behavioral variance. Personality traits are known to have a stronger impact on behavior in settings where appropriate activity is not clearly defined (Stewart, Carson, & Cardy, 1996). In a recent paper, Fischbach and Zapf (2003) proposed that personality traits are crucial to the redefinition process of emotion work because of the leeway of emotion work interpretation. However, the empirically tested effects of personality traits on the redefinition process of emotion work revealed that the influence of personality existed but was less than expected for the redefinition of the work role within a single occupational group. Nevertheless, the body of research on this aspect merits the inclusion of personality as a personal determinant factor influencing the emotion work task redefinition process. The RS Model lists personality as the third of four personal determinants affecting the emotion work task redefinition process.

Gender is also a widely-accepted integral aspect of emotion work (Grandey, 2000; Hochschild, 1983; Schaubroeck & Jones, 2000) and is therefore included in the RS Model. However, it is doubtful whether theoretical conceptualization and associated empirical support for hypotheses regarding gender in the context of the emotion work redefinition process exists.

Presumably, gender influences the redefinition process of an external task, as gender-biased differences in needs and wishes are often considered in the gender and work literature, including the Hackman claim that idiosyncratic needs and

wishes affect task redefinition. It has been proposed that aspects associated with positive emotional behavior such as friendliness and agreeableness are prototypically female; and that aspects of competitiveness and enforcement behavior are prototypically male (Bakan, 1966; Eagly, 1987; Powell, 1999). One explanation offered in the literature for gender differences is that prototypic male and female work role perception and work role behavior arise from the traditional distribution of men and women in specific occupational roles such as physician (prototypically a male work role) vs. nurse (prototypically a female work role). The RS Model lists gender as the fourth of four personal determinants affecting the emotion work task redefinition process. In this instance, it is proposed that personality and gender relate to the idiosyncratic needs, values, attitudes, etc. variable of the Hackman framework personal determinants component. The RS Model lists personality and gender as two of the four personal determinants affecting the emotion work task redefinition process.

It is proposed that personality and gender as personal determinants influence the redefinition process of emotion work (Proposition 9).

2.4 Strategies of emotion work

The second part of the RS Model uses the heading Emotion Work—Strategy to group emotion work conceptualizations dealing with self-regulation processes and variables leading to actual task behavior, emulating the second part of Hackman's (1969, 1970) framework (Figure 1, p. 9). Accordingly, the present study places the Hochschild (1983) claim that several self-regulation processes are typically triggered by emotion work demands in this RS Model Emotion

Work—Strategy stage comprised of a self-regulation process, task behavior, and feedback of task behavior (Figure 2, p. 37). Hochschild focused on these self- regulation processes as necessary to comply with emotion work job demands.

Hochschild defined emotional labor as, "the management of feeling to create a publicly observable facial and bodily display" (p. 7). In cases where the actual emotions of a service worker toward a customer/client are not in line with the display rule (how one should interact with clients), Hochschild proposed that there are two strategies used in order to fulfill the display rules: either surface acting by which only the emotional expression is manipulated in order to fulfill the job demands; or active deep acting by which the feelings of the service worker are actively manipulated in order to fulfill the job demands. The RS Model (Figure 2) lists surface acting and deep acting as two of four strategies evoked by emotional demands involved in emotion work. Ashforth and colleagues (Ashforth & Humphrey, 1993; Ashforth & Tomiuk, 2000) described the phenomenon of authenticity in service where a required emotion is spontaneously felt and displayed by the service worker—what is referred to in the literature as surface authenticity—and what was termed by Hochschild as passive deep acting and later termed automatic emotion regulation by Zapf (2002). Surface authenticity, passive deep acting, and automatic emotion regulation are referred to in this study as automatic regulation—listed in the RS Model as the third of four strategies evoked by emotional demands involved in emotion work. Rafaeli and Sutton (1987) defined a fourth emotion work strategy where the employee does not express the job-required emotions and termed this option emotional deviance. Emotional deviance is included in the RS Model as the fourth emotion work strategy. Grandey (1998) and Kruml and Geddes (1998)

distinguished two high correlating dimensions of emotion work strategies:

emotional dissonance and emotional effort. Emotional dissonance refers to Hochschild's concepts of surface acting and active deep acting; and emotional effort refers to the degree to which employees actively try to change their inner feelings in order to match the feelings they are expected to express. The notion of emotional dissonance, high correlating with the concept of emotional effort, is applied in several studies (Brotheridge & Grandey, 2002; Brotheridge & Lee, 2002; Grandey, 2000) and positive effects of effort on burnout are shown.

This leads to the proposition that emotional demands affect self- regulation efforts (Proposition 10).

Furthermore, emotional demands affect actual task behavior (Proposition11).

Organizational determinants of emotion work strategies. In the self- regulation literature (Bandura, 1977, 1986; Kanfer, 1990) and stress literature (Zapf & Semmer, in press), feedback is suggested and empirically tested as an important determinant in the self-regulation/task behavior dynamic. Feedback can be provided by the customer event, where the service worker learns how task performance strategies either promote or hinder goals ("am I successful with this strategy or should I change my strategy to reach my goal?"). The RS Model lists the organizational determinants of customer event feedback as a variable affecting the self-regulation strategies leading to emotion work task behavior.

The recursive arrow from task behavior to self-regulation (Figure 2, p. 37, RS Model) illustrates the feedback influence of task-behavior in the self-regulation process. Supervisors and colleagues can also be sources of feedback, supporting

service-worker development of more efficient and less stressful emotion work strategies. According to action theory and stress research (Frese & Zapf, 1994;

Hacker, 1973, 1998; Zapf, 2002), job autonomy (leeway in decision-making), and supervisor- and coworker support as well as their feedback promote performance of the service worker. Thus, job autonomy, supervisor- and coworker support, and feedback are predicted to have moderator effects in the relationship between emotional dissonance and negative emotion work consequences. The RS Model lists the organizational determinants of inner organizational job autonomy, supervisor-, and coworker support as well as feedback as variables affecting the self-regulation strategies leading to emotion work task behavior.

It is proposed that organizational determinants of customer event feedback and inner organizational extents of job autonomy, supervisor and coworker support, including feedback, affect self-regulation efforts and task behavior of emotion work (Proposition 12).

Personal determinants of emotion work strategies. In her paper dealing with emotion regulation in the workplace, Grandey (2000) suggested that emotional expressivity and emotional intelligence (presumed in the present study to be components of emotional competence), positive/negative affectivity (widely accepted as personality factors), and gender are included in aspects that influence self-regulation. Emotional competences such as masking true feelings, regulating emotions in self and others, and use of emotions in persuading customers/clients are involved in the emotion work strategy process and affect both the success of this process and employee consequences. The effect of personality (particularly positive or negative affectivity) on negative individual

consequences might be that persons high in negative affectivity respond more intensely to negative events (Weiss & Cropanzano, 1996) and therefore have a greater probability of suffering burnout. In the empirical study of emotion work/emotional labor and burnout by Brotheridge and Grandey (2002), negative affectivity was positively correlated to all three of the burnout dimensions (emotional exhaustion, depersonalization and reduced personal accomplishment).

Grandey (2000) also discussed that women might be more concerned with getting along whereas men might be more concerned with getting by, suggesting different gender-biased emotion work strategies. Influence of personal determinants on emotion work strategies were recently tested by a diary study of Totterdell and Holman (2003), who found that emotional intelligence (measured by a scale of Schutte et al., 1998) was only weakly positively related to deep acting strategies, and that women engaged in both higher levels of negative affect regulation and the faking of emotions (surface acting strategies). Performance, measured by self-description scales (good vs. bad service), was positively related to deep acting strategies but not to surface acting strategies. Emotional exhaustion consequence was positively related to surface acting strategies but not to deep acting strategies. In the self-regulation literature (Bandura, 1977, 1986;

Kanfer, 1990) and stress literature (Jex & Bliese, 1999; Jex, Bliese, Buzzell, &

Primeau, 2001; May, Schwoerer, Reed, & Potter, 1997; Zapf & Semmer, in press), self-efficacy is suggested and empirically tested as a further important personal determinant in the self-regulation process and a moderator in the relationship of the stressor emotional dissonance and negative emotion work consequences in the RS Model. Self-efficacy as a personal determinant seems to exert positive emotion work consequences as self-efficacy influences cognitive

appraisal differences in resources ("am I okay or in trouble?" and "what can I do about this stressful event?") according to the transactional stress model (Lazarus

& Folkman, 1984) and demand resource models (Frese & Zapf, 1994; Karasek, 1979; Oesterreich & Volpert, 1999). The four personal determinant variables discussed in this section—emotional competence, personality, gender, and self- efficacy—are listed as personal determinant components in the strategy stage (second part) of the RS Model.

It is proposed that personal determinants of emotional competence, personality, gender and self-efficacy affect self-regulation efforts and task behavior of emotion work (Proposition 13).

2.5 Consequences of emotion work

This study proposes that several emotion work consequences need to be considered: employee consequences (negative or positive, short-term and/or long-term); customer satisfaction; and inner organizational success (Figure 2, p.

37, RS Model). Once again emulating the Hackman (1969, 1970) input-strategy- outcome framework (Figure 1, p. 9), the RS Model uses the consequence heading to group propositions and findings in the emotion work literature regarding positive and negative consequences of emotion work.

Employee consequences were a pivotal claim of Hochschild's book (1983). She described various negative psychological consequences of emotion work and maintained that showing emotions not felt at the moment (emotional dissonance) leads to the alienation of one's feelings, negative affective states, and eventually causes long-term psychological ill health. Most empirical studies have analyzed relationships between aspects of emotion work and long-term

consequence of emotion work, especially the burnout dimensions (emotional exhaustion, depersonalization and reduced personal accomplishment) and have found correlations between emotional dissonance and emotional exhaustion (Abraham, 1998; Brotheridge & Lee, 1998; Grandey, 1998; Kruml & Geddes, 1998; Morris & Feldman, 1997; Nerdinger & Röper, 1999; Zapf et al., 1999, 2001), as well as correlations between emotional dissonance and depersonalization (Kruml & Geddes, 1998; Zapf et al., 1999). Correlations have also been found between emotional dissonance and reduced personal accomplishment (Abraham, 1998; Morris & Feldman, 1997; Zapf et al., 1999), supporting Hochschild's view that emotion work is likely to have negative employee consequences. This study expands the concept of emotion work and attempts to present it as a multi-dimensional construct. This inclusive approach allows the incorporation of positive correlations that have been found between several aspects of emotion work and job satisfaction (Adelmann, 1995; Morris &

Feldman, 1997; Rutter & Fielding, 1988; Wharton, 1993; Zapf et al., 1999) as a long-term emotion work consequence and corresponding positive affective states as an immediate consequence of emotional regulation and task performance.

In the context of emotion work where the service encounter can be viewed as intrinsic to the job, the criteria for the assessment of task behavior is service quality and includes the quality of the interaction (e.g., SERVQUAL instrument of Parasuraman et al., 1988). In the RS Model it is proposed that an external task redefined to an internal task leads to self-regulation and actual task behavior describable in this example as good vs. bad service quality, and this in turn leads to the inner organizational success or failure and customer satisfaction or dissatisfaction. The RS Model lists organizational consequences as including

consequences impacting the employee, customer satisfaction, and inner organizational success.

This leads to the proposition that employee consequences, customer satisfaction, and inner organizational success are consequences of emotion work—depending on the external task, task redefinition, internal task, self-regulation strategy, and actual task behavior (Proposition 14).

2.6 Summary of the RS Model

The Redefinition Self-Regulation Model of Emotion Work (RS Model) attempts to integrate diverse emotion work conceptualizations and empirical research, applying an organizational focus to emotion work. This organizational focus is especially useful in demonstrating that the emotion work job stressor, emotional dissonance, can be understood as an external task and is subject to the same determinants, strategies, and consequences as any other external task.

The first part of the RS Model entitled "Emotion Work—Job Demand"

attempts to illustrate that emotion work job requirements (displays of positive emotions, negative emotions, neutrality, sensitivity-, and sympathy requirements) and the emotion work job stressor, emotional dissonance, are external tasks which are redefined into internal tasks. These emotion work external tasks are defined and influenced by organizational determinants (occupational, inner organizational, and customer event) and this body of organizational determinants, working together with personal determinants (professional identity, emotional competence, personality, and gender), are involved in the redefinition process that redefines external tasks into internal tasks.

The second part of the RS Model is entitled "Emotion Work—Strategy"

and is characterized by strategies (surface acting, deep acting, automatic regulation, or emotional deviance) evoked by emotional demand. These strategies require different degrees of effort of self-regulation and actual task behavior. Self-regulation is influenced by organizational determinants of customer event feedback and inner organizational extents of job autonomy, supervisor- and coworker support including their feedback. Self-regulation is also influenced by personal determinants of emotional competence, personality, gender, and self-efficacy. The self-regulation process leads to actual task behavior—providing recurrent customer event feedback that influences the self- regulation process.

Emotion work task behavior leads to the third and final part of the RS Model entitled "Consequences." Three categories of organizational consequences are: various employee consequences, various levels of customer satisfaction, and various levels of inner organizational success. Employee consequences are characterized by ranges of negative or positive, short-term and/or long-term, and are listed as negative affect, performance, burnout, positive affect, and job satisfaction. Figure 2 illustrates the Redefinition Self-Regulation Model of Emotion Work.